Slashdot Mirror
IBM's New Processors To Exceed 5Ghz
Jordin Normisky writes to mention the news, via ZDNet Asia, that IBM's new Power6 processor will be unveiled next month at a conference in San Francisco. They're also planning to announce a second-generation Cell, both of which are expected to run faster than 5GHz. From the article: "In addition, the [Power6] chip 'consumes under 100 watts in power-sensitive applications,' a power range comparable to mainstream 95-watt AMD Opteron chips and 80-watt Intel Xeon chips. Power6 has 700 million transistors and measures 341 square millimeters, according to the program. The smaller that a chip's surface area is, the more that can be carved out of a single silicon wafer, reducing per-chip manufacturing costs and therefore making a computer more competitive. Power6, like the second-generation Cell, is built with a manufacturing process with 65-nanometer circuitry elements, letting more electronics be squeezed onto a given surface area. "
I thought we had finally advanced past the "higher clockspeed = more better" stage...
- Toby
Usually from the bell-end of Apple. I wonder if IBM's fab plants can cash the check their PR department writes.
Do you even lift?
These aren't the 'roids you're looking for.
If I owned an as/400, i/z series server, maybe this would be exciting news as a future upgrade.
But do they achieve a comparable amount of work per cycle?
--
Wi-Fizzle Research
Censorship is obscene. Patriotism is bigotry. Faith is a vice. Slashdot 2.0 sucks.
The smaller that a chip's surface area is, the more that can be carved out of a single silicon wafer, reducing per-chip manufacturing costs and therefore making a computer more competitive. Power6, like the second-generation Cell, is built with a manufacturing process with 65-nanometer circuitry elements, letting more electronics be squeezed onto a given surface area.
The cost of making chips, by far, is the R&D cost. The "first" chip costs hundreds of millions to make. Once the "first chip" is made the margin cost is VERY low. Beyond recovering R&D costs....the rest is just distribution channel costs....then....PROFIT!They would get bragging rights with 45nm. 65nm is so old that even AMD has 65nm chips now.
Heck philips/motorola I believe have been producing 65nm microcontrollers, and samsung is producing 50nm flash chips.
And 5GHz should not be difficult considering it doesnt have the x86 overhead, is more RISC and that generally PPC has a simpler core. I'll be interested if it comes with quad cores or more.
"Give orange me give eat orange me eat orange give me eat orange give me you." -Nim Chimpsky
...undisturbed 802.11a networking!
Tsunami -- You can't bring a good wave down!
They're also planning to announce a second-generation Cell, both of which are expected to run faster than 5GHz.
Why don't they seem to be making any kind of performance comparisons? Talking about physical size, power consumption as compared to intel & AMD are great, but it seems weird that there's no mention of real-world performance against those same competitors. Even a rough estimate would be interesting.
The theory of relativity doesn't work right in Arkansas.
In your face, Steve Jobs!
Boy, Howdy! are you out of the loop. I work on those suckers and believe you me, the chip cost is not trivial.
Do the math: the cost of a 300 mm wafer in a 65 nm process runs well over $5000 (how much is a Deep Dark Secret.) Ignoring geometric yield loss, that's about 70,000 mm of potential dice per. If one chip is 350 square mm, you're getting about 200 per wafer, or $25 per chip fab cost. Yield drops off steeply with size (think in terms of losing ten to twenty dice per wafer, regardless of die size) and that adds into the fab cost too.
That's bare minimum, assuming there aren't any bad lots etc. It adds up fast.
Lacking <sarcasm> tags,
It would be ludicrous, but Kutaragi's talked before about never reducing the price of the PS3 but instead upgrading it with more memory, bigger hard drives, etc. It would be pretty damned amusing if, a year and a half after PS3 launch, instead of cutting prices with a new easier to produce Cell and Blu-ray they upgraded the PS3 with the Cell2(and hosed everyone who'd already bought one). This would be so stupid and arrogant that it's only plausible because it's Sony.
Move back? They were never on them. POWER6 != powerpc (though they are similar in more ways than not).
I think Apple is perfectly happy with the Intel move at this point. One of the reasons for the migration (if you can get past Jobs' reality distortion field of blah blah per watt or whatever) was that IBM wasn't able to keep up with demand, either with getting the speeds up, or with delivering the slow crappy ones they already had.
/* oops I accidentally made a comment, sorry */
First of all, switch to a Power6 based architecture is not something you simple do. It takes a LOT of effort in writing the OS to function on the new architecture, not to mention all the work by developers to make their programs function on it as well. Second, Apple didn't choose Intel because they were the "best at the moment" uP supplier. They chose Intel because Apples felt they had a better future than the PowerPC line. So, even if someone, like Power6, does poke their head above Intel/x86 in performance, Apple is content that Intel will surpass them and continue producing good CPU's. Apple did not switch to x86 based processors lightly.
Space for rent, inquire within
"NOW you tell us?"
-Steve Jobs
we've moved past the megahertz myth. we're stuck on the gigahertz myth now.
In the world of technology a promise of more/better performance counts as much as a drunken "I love you." One reason why Apple jumped from PPC is that IBM failed to deliver a 3.0 Ghz chip within a reasonable time frame (in the PPC970 series) and completely failed on delivering a laptop chip. Believe it when you actually see shipping servers.
Leave the gun, take the cannoli -- Clemenza, The Godfather
It was never about performance per se -- there are plenty of faster things out there than the Core 2 Duo. IBM will be happy to sell you some of them, as will Sun or Fujitsu. Or Cray. All for the low price of $600k a machine.
The issue is that IBM makes supercomputers, and Motorola makes cellphones, and they design their chips accordingly. Apple, making neither of these things, couldn't persuade either of them to make a low-power, fast, cheap CPU useful for a laptop and continue updating it with such a small market. Intel, on the other hand, spends most of their engineering effort trying to solve exactly this problem, and so has its business interests aligned with Apple's, as opposed to IBM or Motorola, who didn't really care about them at all, and would happily spend their R&D money on designing things like this chip instead of making a G5 that would fit in a laptop.
That's all right, I still find myself stumped by analog processors, like the valve body in a GM 700R automatic transmission. *shudder*
Anyway, here goes:
Basically they take a tiny wafer-thin piece of silicon, use chemical to scrape out millions of little transistor shapes onto its surface, and strap a buckin' bronco of a clock crystal on it that shakes it like a salt shaker, or like jello jigglers on free-based cocaine.
Thusly, the outrageous oscillating action of Mr. crystal causes the tiny transistorized citizens to go into a tizzy, but they're all right because they're not hollow and fragile like vacuum tubes, so they get all busy and start swinging their logic gates open and shut kind of like an electron square dance.
The speed that the crystal is eventually set at is the maximum speed at which the transistors can go about their daily lives, such as munching on electrons and crapping them out, organising meetings, forming PTA (parent-transistor association) clubs, and dipping into the cache without generally spontaneously combusting or reverting back to silicon amoebas. Each time the crystal wiggles its booty constitutes one clock cycle, and the number of operations per cycle varies based on the processor and the types of instructions the poor transistors must labor over.
Clock speed has been historically limited by various things, including level 2 cache memory timings (remember this stuff has been running at CPU speed for about 5 years now!), motherboard design, pipeline depth, heat dissipation, ALU and/or FPU limitations, or even the leakage the P4 was subject to at 4GHz+ clockspeeds. Right now I believe it's the fault of lazy hardware engineers at Intel and even AMD. Dual-core and quad-core is an easy out for expensive, fast-sounding hardware just like it is for the video card market right now, and the burden of performance improvement has been shifted to software engineers (reducing bloat, multithreading applications, both of which can only go so far). IBM is hopefully going to prove this with a higher-clocked POWER chip or two that maintains the efficiency they have a reputation for, although we may never see the return of single-core CPUs for performance systems.
I may make you feel, but I can't make you think.
With 32MB of cache, hopefully cache misses won't be too infrequent. IBM, as well as being the first to market with dual- and quad-core, were first to market with SMT as well. The nice thing about SMT is that when you get a cache miss, you can just give the other thread a bit more time to run. With enough contexts (and a high enough degree of parallelism) cache misses become much less important. This is something the T1 does particularly well.
I am TheRaven on Soylent News
IBM does not give a heck to Desktop market unless you are calling them about 10.000 terminal running Enterprise Big Iron monster and they may even suggest you buy Dell terminals/PCs if it fits their project better. What matters to them is the mainframe, technologies used, software used and the entire consulting to keep such business up.
Motorola/Freescale lives happily in embedded processor market and telecoms market too.
I guess such stories should have "power-not-powerPC department" tag.
Also, yes , our great leader/prophet whatever was right switching to Intel/x86 because of above reasons. Both companies tries to stay away from Desktop market and they won't be bothered by ridiculous 3Ghz PPC G5 (a STRIPPED DOWN POWER4) Apple fanboys. Apple can't effect those decisions by their current market share. If it goes back to great 50% 50% marketshare values, they can demand anything of course.
(Happily written from a 33C/92F running Quad G5)
Sorry, but these annoucements arent much more optimistic than the ones that were made before the launch of the G5.
Lets see IBM actually roll out those babies, and look what yields they get, how cool they really run and in what ways the design has suffered to allow them to reach that kind of clockspeeds.
HI O WISE PRINCE. WHT TOOK U SO DAM LONG?
I haven't checked the information yet, but here's an abstract on the rest, found through google:
c essor-Trashes-Competition-with-6-GHz-17765.shtml1 606194731
The Power6 processor will run between 4GHz and 5GHz and it has been proven to chew away data at a speed of 6GHz in the lab.
IBM see things a little differently and they decided to raise the frequency in both cores of the processor.
For high-end models, four POWER6 MPUs will be packaged in a single multi-chip module, along with four L3 victim caches, each 32MB.
On the management side, IBM is also improving their virtualization capabilities in the POWER6. In particular products, a single processor may be able to host 2-300 virtual instances, although theoretically up to 1024 VMs are possible. Memory partitioning and migration have been added as well, which reduces system down time for repairs.
IBM is claiming a factor of two performance increase, which would be consistent with the vastly higher clockspeeds and increases in raw system bandwidth.
IBM's roadmaps currently include the POWER6+, which is presumably a 45nm derivative product. Judging by past practices, the POWER6+ will debut in the second half of 2008, probably just in time to dash the hopes of rivals.
The Power and PowerPC lines will grow one step closer together with Power6, which incorporates the AltiVec instruction set that speeds up many multimedia tasks. AltiVec, also known as VMX, increases efficiency by letting a single processing instruction be applied to multiple data elements. That's helpful for video and audio tasks on desktop machines, but servers will benefit as well in, for example, high-performance computing tasks such as genetic data processing, McCredie said
Where Power5 can transfer data on and off the chip at a rate of 150 gigabytes per second, Power6 can do so at 300GBps, McCredie said.
Oh, and it is also good for BCD's (binary coded decimals) which obviously points to the expected customers (high end financial firms, presumably).
Sources:
http://news.softpedia.com/news/New-Power6-IBM-Pro
http://realworldtech.com/page.cfm?ArticleID=RWT10
http://news.zdnet.com/2100-9584_22-6124451.html
5 Ghz is enough speed for everyone.